zation of a given stock of facilities (Chapter 5). He was specifically concerned with the average and marginal costs of short-run increases in the rate of case-flow. The marginal cost of a case was found to be almost 21% of average cost, calculated at the mean values of the independent variables. Feldstein also calculated the MC/AC ratios for seven individual case-mix categories. The MC/AC ratios, by catetgory, are: general medical, 0.17; pediatrics, 0.26; surgery, 0.21; ear-nosethroat, 0.47; gynecology, -0.03; obstetrics, 0.47; and others, 0.40. Feldstein also found a serious underutilization of the stock of hospital facilities in the short-run.

After evaluating the literature, with special emphasis on the above three studies, and with the assistance of the Bureau of Health Services Research and Evaluation of the Department of Health, Education and Welfare, a marginal cost to average cost factor of 40% was included in the Phase IV regulations. If this adjustment were used from zero change in admissions, it would mean that an institution that was increasing in volume would be allowed only 40% of average cost and charge for each extra admission. The hospital that was declining in volume would be allowed to keep 60% of the average cost and charge of each lost admission, and recover the difference from the remaining patients.

There are, however, statistical indications that while the 40/60 ratio is a reasonable average, when volume of admissions increases the marginal cost may actually be somewhat higher, i.e., it costs more than 40% to service the increased admissions; and that when volume is declining the fixed costs may be higher, 1.e., a hospital cannot reduce its cost per case by 40%. For this

30 The American Hospital Association has been concerned with the 60-40 split of costs. They presented data that show that on the up-side, marginal costs are 80% of average costs and on the down-side, they are only 10-20% of average cost.

reason, and to reduce the need for hospitals to make charge adjustments for small changes in volume, a "zone of no adjustment" was created.

ZONE OF NO ADJUSTMENT

The imposition of the Phase IV volume adjustment is delayed for most hospitals, with the extent of the delay dependent on their size. For the first 2% increase in admissions for large hospitals (and 4% increase for small 21) the 7.5% allowance is permitted over last year's per admission averages, with no adjustment required. That is to say, the hospital is allowed to assume that all costs are variable.

The large hospital that declines up to 5% in admissions (10% for small hospitals) is allowed to assume that all costs are fixed. That is to say, the hospital may keep its total budget at 107.5% of the last year's budget. Once volume changes go beyond these limits, the volume adjustment begins. The equations for calculating the percentage limits for each change in volume appear in the appendix. Much has been made of the marginal cost assumption and the potential harshness of the volume adjustment. The effect of the zone, however, is to postpone the volume adjustment and to liberalize the marginal cost assumption. Table 9 shows, for both large and small hospitals, the actual marginal cost assumption of the regulations, depending upon the change in volume that is actually experienced. A further modification in the volume adjustment formula is a limit on the required reduction in price for volume increases such that a hospital could impose at least a 3% increase in cost per admission. On the down-side a limit of 20% was imposed on the allowable cost increase per admission. These are shown in Table 9.

"A small hospital is defined as one with less than $2.5 million in budget or less than 4,000 admissions.

CASE-MIX ADJUSTMENT

There has also been much concern about the flexibility of the Phase IV system when a hospital's patient-mix changes. As many hospital services are shifted to out-patient settings, and as more illnesses become treatable, hospital patient-mix will become generally more complex, and perhaps also more expensive. There can also be shifts among institutions, with certain hospitals becoming referral centers for difficult cases.

The Phase IV controls include adjustment of revenues and costs that recognizes the increased costs of a more complex patientmix. It is not part of the basic control system, and is placed in force at the option of the hospital. It will not be used to penalize hospitals that shift to less complex patients and is a self-executing adjustment.

The patient-mix adjustment procedure relies on either of two standard patient allocation systems. In addition, a hospital may use another allocation system with prior COLC approval. The procedure to allocate charges and expenses works in the following way. The current year's distribution of patients is determined by the category list below, and last year's total admissions are restated in these proportions.

Allocation of Patients

1. Medical, Surgical, Obstetrics, Pediatrics, Psychiatric; OR

2. H-ICDA diagnostic codes representing at least 85% of the hospital's admissions, the balance being included in a residual “other” category.

A restated "total charges" for the last year is then calculated using the actual charge per case (in each category) multiplied by the restated number of admissions in that category, with all the results added. The restated total budget is then divided by the actual total admissions to get a restated average charge per admission for the last year. The percent by which this amount exceeds

last year's actual average charge per admission is the increment which could be added to the basic control allowances for charges and expenses. There is, however, a potential bias in the assumption that expenses due to patient-mix changes will change by the same percentage as the charges. The more likely case is that the difference in costs would be greater than the difference in charges. This is due to the fact that charges for high cost items are often redistributed to room and board rates. Patients do not pay $800 per day for coronary care. They pay perhaps $250, with all the other patients absorbing $1-2 in their room rates. Therefore charge ratios might not reflect cost differences. If a hospital felt that the data were indeed not reflective of a particular situation, it could request an exception.

INCENTIVES UNDER PHASE IV

As in any system of control based on a particular unit of output, the incentive is to increase the volume of that unit. That is true of the Phase IV system as it was under Phase II. But Phase IV includes three features to minimize such undesirable incentives:

(1) It is more difficult to create an unnecessary admission than to extend each hospital stay one extra day or order some extra tests "just in case”.

(2) The simple outpatient limitation (with no intensity limitation) does not inhibit the trend toward more services being performed on an outpatient basis.

(3) Finally, and most importantly, the hospital that increases its volume of admissions beyond the zone is required to recognize that all costs are not variable, and that some economies of scale should be realized. In contrast, the Phase II system allowed free rein to increase costs and charges one-for-one with volume increases.

The Phase II system worked in such a way as to have undesirable effects on the hospital that had declining length of stay or declin